Colorectal adenomas (CRAs) are the benign precursors of most cases of colorectal cancer (CRC), a leading cause of cancer deaths; advanced CRAs are those most likely to progress to CRC. Although most CRAs can be removed at colonoscopy, 20-50% of individuals undergoing repeat colonoscopy 3-5 years later have metachronous, i.e. new, CRAs. Furthermore, a proportion of CRCs are not preventable through screening colonoscopy, necessitating alternative preventive strategies, such as chemoprevention. When given as a supplement, selenium (Se), a trace dietary mineral that is incorporated into specialized selenoproteins was previously shown to protect against CRC and prevalent CRAs as secondary endpoints in a non-melanoma skin cancer chemoprevention trial. Chemopreventive agents for CRC can be assessed in randomized controlled trials (RCTs) comparing metachronous CRA rates between intervention and placebo groups with results of effects on CRA recognized by the FDA as a CRC surrogate. No RCTs of Se supplements for CRC chemoprevention or for any other preneoplastic condition as the primary endpoint, have been reported. This application is for 3 years funding to complete The Selenium Trial (SeT); an RCT of Se, 200 5g daily as selenized yeast, for which randomization is completed. The primary study hypothesis is that treatment with Se yeast for 3 to 5 years will reduce the rate of metachronous CRAs without associated serious toxicities. Funding is requested to follow remaining SeT participants through study completion for all 1,800 participants in early 2014, and for data analysis and reporting. A major new study aim has been added to address the suggestion in recent published studies that excessive Se exposure may increase risk for type 2 diabetes (T2D). Although a mechanism for Se-related T2D in human is obscure, animal studies suggest that chronic activation of glutathione peroxidase 1 (GPX1), a potent anti-oxidant selenoprotein, may lead to the development of hyperinsulinemia, hyperglycemia, insulin resistance, and obesity. Therefore, the question of whether or not Se supplementation contributes to T2D has important public health implications. In the SeT cohort, we propose two approaches to studying the effects of Se on T2D, and insulin sensitivity and secretion: an epidemiological approach using the entire cohort; and assessment of the effect of Se supplementation on insulin secretion and sensitivity at the individual level using a modified oral glucose tolerance test in a sub- cohort of 300 subjects (150 on placebo and 150 on Se yeast). In secondary analyses, we will investigate whether or not baseline Se levels, genetic background, and/or use of low-dose aspirin modify Se effects on CRA occurrence or risk for T2D.